> ## Documentation Index
> Fetch the complete documentation index at: https://se7en.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# 训练

> 预训练、后训练、分布式训练与训练基础设施项目

## RL 训练

| 项目                                                          | 简介                                                                                                                                                       | 参考资料                                                                                                                                                                                                                            |
| ----------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| [Accio-Lab/Dressage](https://github.com/Accio-Lab/Dressage) | 基于 slime 的 Agentic RL 训练框架，通过统一的 Proxy 与 Paddock 层连接 Agent Rollout、Sandbox 工具执行、Token 级轨迹记录和训练样本转换，并支持 Python 白盒工具循环与 Codex、Claude Code 等 HTTP 黑盒 Agent。 | [快速开始](https://github.com/Accio-Lab/Dressage/blob/main/docs/quickstart.md) · [训练机制](https://github.com/Accio-Lab/Dressage/blob/main/docs/training.md) · [训练配方](https://github.com/Accio-Lab/Dressage/blob/main/docs/recipes.md) |